omicverse-single-cell-batch-integration

Automate batch integration of preprocessed single-cell AnnData across multiple backends.

13|2|Updated Apr 5, 2026
One-click install
npx skills add https://github.com/Starlitnightly/omicverse-skills --skill omicverse-single-cell-batch-integration
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: omicverse-single-cell-batch-integration
Source: https://github.com/Starlitnightly/omicverse-skills/tree/main/src/omicverse_skills/skills/single-cell-batch-integration
Command: npx skills add https://github.com/Starlitnightly/omicverse-skills --skill omicverse-single-cell-batch-integration

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires anndata, numpy, pandas, omicverse, scib_metrics, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

Automates batch integration for preprocessed single-cell AnnData across multiple backends, enabling side-by-side comparison and benchmarking.

Core Features & Use Cases

  • Supports integration backends Harmony, Combat, Scanorama, scVI, CellANOVA, and Concord from a single callable workflow, streamlining technology comparisons.
  • Enables benchmarking by producing both integrated embeddings and non-integrated references for evaluation across methods.
  • Real-world use: compare batch correction strategies on a preprocessed AnnData with a batch column to select the best backend for downstream analysis.

Quick Start

Run the preprocessed AnnData through the batch-integration skill to compare Harmony, Combat, and ScVI backends and optionally run the Benchmarker.

Frequently Asked Questions about omicverse-single-cell-batch-integration

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I perform single-cell batch integration on preprocessed AnnData?

Single-cell batch integration applies to preprocessed AnnData with a batch column, utilizing multiple backends like harmony, combat, scanorama, and scVI to generate integrated embeddings for downstream analysis.

Can I compare different batch correction methods like harmony and scVI?

Yes, you can compare batch correction methods like harmony and scVI side-by-side. The workflow supports multiple backends and optional benchmarking using scib_metrics to evaluate and select the best strategy.

What inputs do I need to run scanorama or scVI batch integration?

Running scanorama or scVI batch integration requires explicit inputs for backend selection, PCA settings, and backend-specific prerequisites on your preprocessed AnnData object to return integrated embeddings.

Does this batch integration workflow support CellANOVA and Concord?

Yes, the batch integration workflow supports CellANOVA and Concord backends alongside harmony, combat, scanorama, and scVI, returning integrated embeddings or models as appropriate for the selected method.

Why do I need a non-integrated reference for single-cell batch benchmarking?

A non-integrated reference is needed for single-cell batch benchmarking to evaluate the effectiveness of integrated embeddings against the original data using scib_metrics for objective method comparison.

What are the limitations of using combat for single-cell batch integration?

Combat provides batch integration but requires explicit PCA settings and backend-specific prerequisites. Users must provide a preprocessed AnnData with a batch column to successfully generate integrated embeddings.